Abstract Currently, the Draize live rabbit eye test is used to distinguish consumer products and materials that are safe from those that are hazardous and can cause reversible or irreversible ocular damage. However, new laws are currently being, or are planned to be, implemented in California, the United States, and the European Union to ban the use of animals in ocular irritation testing. Despite this fact, there is currently no single nonanimal test or combination of nonanimal tests that can accurately classify ocular irritants. It is of particular concern that all nonanimal tests generally overpredict the least damaging class of ocular irritants (EPA IV and GHS nonclassified materials) and cannot determine (or determine with low sensitivity) the most extreme class (EPA I and GHS 1 irreversible/corrosive materials) for which they can have sensitivity rates as low as 50%. More importantly, none of the currently validated alternative tests can detect either reversible irritants (EPA II/III and GHS 2A/B) or those materials that cause irritation that lasts from 7 to 21 days but then clears without permanent corneal damage. This target classification represents a critical UNMET NEED for correctly informing and advising consumers and manufacturers about product safety. We have recently developed and tested a new method?the in vitro depth of injury eye irritation test (IVD EIT) to detect reversible ocular irritants with high sensitivity and specificity. The purpose of this application is to further develop the IVD EIT with the long- term goal of providing a service and/or product kit for the accurate classification of reversible ocular irritants to meet this UNMET NEED as part of a nonanimal ocular irritation testing strategy. This goal will be achieved by accomplishing the following specific aims: Aim 1 (2 months): Assess the repeatability and reproducibility of the IVD EIT by performing a repeat-measures experiment of a four-chemical test set composed of EPA I?IV chemicals to determine the number of corneas and positive controls needed to achieve high repeatability. Aim 2 (4 months): Perform a prevalidation test to assess the ability of the IVD EIT to categorize 30 test substances according to GHS and EPA classifications and further refine the prediction model. Aim 2 will specifically determine the accuracy, sensitivity, and specificity of the IVD EIT and identify any application domain limitations (the range of irritant categories and chemical/physical properties detected). If the results from Aim 2 validate and demonstrate the commercial and regulatory utility of the IVD EIT as defined by the ability to detect reversible irritants, we would then propose a phase II SBIR project. 1